| Age | Commit message (Collapse) | Author |
|---|
|
Controls Qualcomm's SS PHY 1.0.0 implemented on various SoCs on both the
20nm and 28nm process nodes.
Based on Sriharsha Allenki's <sallenki@codeaurora.org> original code.
[bod: Removed dependency on extcon.
Switched to gpio-usb-conn to handle VBUS On/Off
Switched to usb-role-switch to bind gpio-usb-conn to DWC3]
Signed-off-by: Jorge Ramirez-Ortiz <jorge.ramirez-ortiz@linaro.org>
Cc: Jorge Ramirez-Ortiz <jorge.ramirez.ortiz@gmail.com>
Cc: Sriharsha Allenki's <sallenki@codeaurora.org>
Cc: Andy Gross <agross@kernel.org>
Cc: Bjorn Andersson <bjorn.andersson@linaro.org>
Cc: Kishon Vijay Abraham I <kishon@ti.com>
Cc: Philipp Zabel <p.zabel@pengutronix.de>
Cc: linux-arm-msm@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
Reviewed-by: Philipp Zabel <p.zabel@pengutronix.de>
Tested-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
|
|
Binding description for Qualcomm's Synopsys 1.0.0 SuperSpeed PHY. This PHY
appears in a number of SoCs on various flavors of 20nm and 28nm nodes.
This commit adds information related to the 28nm node only.
Based on Sriharsha Allenki's <sallenki@codeaurora.org> original
definitions.
[bod: converted to yaml format]
Signed-off-by: Jorge Ramirez-Ortiz <jorge.ramirez-ortiz@linaro.org>
Cc: Jorge Ramirez-Ortiz <jorge.ramirez.ortiz@gmail.com>
Cc: Rob Herring <robh+dt@kernel.org>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Bjorn Andersson <bjorn.andersson@linaro.org>
Cc: Jorge Ramirez-Ortiz <jorge.ramirez.ortiz@gmail.com>
Cc: devicetree@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
Reviewed-by: Rob Herring <robh@kernel.org>
Tested-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
|
|
Adds Qualcomm 28nm Hi-Speed USB PHY driver support. This PHY is usually
paired with Synopsys DWC3 USB controllers on Qualcomm SoCs.
The PHY can come in two flavours femtoPHY or picoPHY. This commit adds
support for the femtoPHY with the possibility of extending to the picoPHY
with additional future commits. Both PHYs are on a 28 nanometer process
node.
[bod: Updated qcom_snps_hsphy_set_mode to match new method signature
Added disjunct on mode > 0
Removed regulator_set_voltage() in favour of setting floor in dts
Removed 'snps' and from driver name
Extended commit log to mention femtoPHY and picoPHY for future
reference.]
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
Cc: Andy Gross <agross@kernel.org>
Cc: Bjorn Andersson <bjorn.andersson@linaro.org>
Cc: Kishon Vijay Abraham I <kishon@ti.com>
Cc: Philipp Zabel <p.zabel@pengutronix.de>
Cc: Jorge Ramirez-Ortiz <jorge.ramirez.ortiz@gmail.com>
Cc: linux-arm-msm@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
Tested-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
|
|
Adds bindings for Qualcomm's 28 nm USB PHY supporting Low-Speed, Full-Speed
and Hi-Speed USB connectivity on Qualcomm chipsets.
[bod: Converted to YAML. Changed name dropping snps, 28nm components]
Signed-off-by: Sriharsha Allenki <sallenki@codeaurora.org>
Signed-off-by: Anu Ramanathan <anur@codeaurora.org>
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
Reviewed-by: Rob Herring <robh@kernel.org>
Tested-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Cc: Andy Gross <agross@kernel.org>
Cc: Bjorn Andersson <bjorn.andersson@linaro.org>
Cc: Kishon Vijay Abraham I <kishon@ti.com>
Cc: Rob Herring <robh+dt@kernel.org>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Jorge Ramirez-Ortiz <jorge.ramirez.ortiz@gmail.com>
Cc: linux-arm-msm@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
Cc: devicetree@vger.kernel.org
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
|
|
This binding is not used by any driver.
Signed-off-by: Jorge Ramirez-Ortiz <jorge.ramirez-ortiz@linaro.org>
Cc: Jorge Ramirez-Ortiz <jorge.ramirez.ortiz@gmail.com>
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
|
|
Usually the digital and analog phys use the same reference clock,
but some platforms have two separate reference clocks for each of
them, so add another optional clock to support them.
In order to keep the clock names consistent with PHY IP's, change
the da_ref for analog phy and ref clock for digital phy.
Signed-off-by: Chunfeng Yun <chunfeng.yun@mediatek.com>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
|
|
The u3phya_ref clock is already moved into sub-node, and
renamed as ref clock, no used anymore now, so remove it,
this can avoid confusion when support new platforms
Signed-off-by: Chunfeng Yun <chunfeng.yun@mediatek.com>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
|
|
Sometimes the reference clock of USB3 PHY comes from oscillator
directly, and no need refer to a fixed-clock in DTS anymore
if make it optional.
Signed-off-by: Chunfeng Yun <chunfeng.yun@mediatek.com>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
|
|
This is used to tune J-K voltage by internal R (resistance), the
range is [0, 31], the resistance value is about 6.9K ohm for 0,
3.8K ohm for 31, and the step is 1K ohm
Signed-off-by: Chunfeng Yun <chunfeng.yun@mediatek.com>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
|
|
This is used to tune the threshold of disconnect, the index range
is [0, 15], the threshold voltage is about 400mV for 0, 700mV for
15, and the step is 20mV.
Signed-off-by: Chunfeng Yun <chunfeng.yun@mediatek.com>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
|
|
Add three required properties about the address mapping, including
'#address-cells', '#size-cells' and 'ranges'
Signed-off-by: Chunfeng Yun <chunfeng.yun@mediatek.com>
Reviewed-by: Rob Herring <robh@kernel.org>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
|
|
Usually the digital and analog phys use the same reference clock,
but on some platforms, they are separated, so add another optional
clock to support it.
In order to keep the clock names consistent with PHY IP's, use
the da_ref for analog phy and ref clock for digital phy.
Signed-off-by: Chunfeng Yun <chunfeng.yun@mediatek.com>
Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
|
|
The u3phya_ref clock is already moved into sub-node, and
renamed as ref clock, no used anymore now, so remove it
to avoid confusion
Signed-off-by: Chunfeng Yun <chunfeng.yun@mediatek.com>
Reviewed-by: Rob Herring <robh@kernel.org>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
|
|
Make the ref clock optional, then we no need refer to a fixed-clock
in DTS anymore when the clock of USB3 PHY comes from oscillator
directly
Signed-off-by: Chunfeng Yun <chunfeng.yun@mediatek.com>
Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
|
|
Add two optional properties, one for tuning J-K voltage by INTR,
another for disconnect threshold, both of them are related with
connect detection
Signed-off-by: Chunfeng Yun <chunfeng.yun@mediatek.com>
Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
|
|
Arguments are supposed to be ordered high then low.
Signed-off-by: Joe Perches <joe@perches.com>
Reviewed-by: Neil Armstrong <narmstrong@baylibre.com>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
|
|
The PCIe PHY initialization requires the attached device to be present,
which is primarily achieved by the PCI controller driver. So move the
logic from init/exit to power_on/power_off.
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Reviewed-by: John Stultz <john.stultz@linaro.org>
Reviewed-by: Vinod Koul <vkoul@kernel.org>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
|
|
Since this phy is shared by multiple devices including USB and PCIe,
it is necessary to determine which device use this phy.
This patch adds SoC-dependent functions to determine a device using
this phy.
When there is 'socionext,syscon' property in the pcie-phy node,
the driver calls SoC-dependt function instead of checking .has_syscon
in SoC-dependent data. The function configures the system controller
to use phy for PCIe.
Signed-off-by: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
|
|
Add legacy SoC support that needs to manage gio clock and reset and to skip
setting unimplemented phy parameters. This supports Pro5.
This specifies only 1 port use because Pro5 doesn't set it in the power-on
sequence.
Signed-off-by: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
|
|
In case of using default parameters, communication failure might occur
in rare cases. This sets Rx sync mode parameter to avoid the issue.
Signed-off-by: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
|
|
Add legacy SoC support that needs to manage gio clock and reset.
This supports Pro5.
Signed-off-by: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
|
|
Pro5 SoC has same scheme of USB3 ss-phy as Pro4, so the data for Pro5 is
equivalent to Pro4.
Signed-off-by: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
|
|
This adds compatible string for Pro5 SoC that needs to manage gio clock
and reset. And Pro4 SoC uses USB2 PHY instead of USB3 HS-PHY, so this
removes Pro4 description from usb3-hsphy.
Signed-off-by: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
Reviewed-by: Rob Herring <robh@kernel.org>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
|
|
Use devm_platform_ioremap_resource() to simplify the code.
Signed-off-by: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
|
|
The support for the 14nm MSM8996 UFS PHY is currently handled by the
UFS-specific 14nm QMP driver, due to the earlier need for additional
operations beyond the standard PHY API.
Add support for this PHY to the common QMP driver, to allow us to remove
the old driver.
Acked-by: Rob Herring <robh@kernel.org>
Reviewed-by: Vinod Koul <vkoul@kernel.org>
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
|
|
Implement single link subnode support to the phy driver.
Add reset support including PHY reset and individual lane reset.
Signed-off-by: Swapnil Jakhade <sjakhade@cadence.com>
Signed-off-by: Yuti Amonkar <yamonkar@cadence.com>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
|
|
Add platform dependent initialization data for Torrent PHY used in TI's
J721E SoC.
Signed-off-by: Swapnil Jakhade <sjakhade@cadence.com>
Signed-off-by: Yuti Amonkar <yamonkar@cadence.com>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
|
|
Use regmap to read and write DPTX specific PHY registers.
Signed-off-by: Swapnil Jakhade <sjakhade@cadence.com>
Signed-off-by: Yuti Amonkar <yamonkar@cadence.com>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
|
|
Use regmap for accessing Torrent PHY registers. Modify register offsets
as defined in Torrent PHY user guide. Abstract address calculation
using regmap APIs.
Signed-off-by: Swapnil Jakhade <sjakhade@cadence.com>
Signed-off-by: Yuti Amonkar <yamonkar@cadence.com>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
|
|
Add support for PHY configuration APIs. These will mainly reconfigure
link rate, number of lanes, voltage swing and pre-emphasis values.
Signed-off-by: Swapnil Jakhade <sjakhade@cadence.com>
Signed-off-by: Yuti Amonkar <yamonkar@cadence.com>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
|
|
Add configuration functions for 19.2 MHz refclock support.
Add register configurations for SSC support.
Signed-off-by: Swapnil Jakhade <sjakhade@cadence.com>
Signed-off-by: Yuti Amonkar <yamonkar@cadence.com>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
|
|
Add a separate function to set different power state values.
Use uniform polling timeout value. Also check return values
of functions for proper error handling.
Signed-off-by: Swapnil Jakhade <sjakhade@cadence.com>
Signed-off-by: Yuti Amonkar <yamonkar@cadence.com>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
|
|
Add wrapper functions to read, write DisplayPort specific PHY registers to
improve code readability.
Signed-off-by: Swapnil Jakhade <sjakhade@cadence.com>
Signed-off-by: Yuti Amonkar <yamonkar@cadence.com>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
|
|
Add a wrapper function to write Torrent PHY registers to improve
code readability.
Signed-off-by: Swapnil Jakhade <sjakhade@cadence.com>
Signed-off-by: Yuti Amonkar <yamonkar@cadence.com>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
|
|
- Change private data struct cdns_dp_phy to cdns_torrent_phy
- Change module description and registration accordingly
- Generic torrent functions have prefix cdns_torrent_phy_*
- Functions specific to Torrent phy for DisplayPort are prefixed as
cdns_torrent_dp_*
Signed-off-by: Swapnil Jakhade <sjakhade@cadence.com>
Signed-off-by: Yuti Amonkar <yamonkar@cadence.com>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
|
|
Rename Cadence DP PHY driver from phy-cadence-dp to phy-cadence-torrent
to make it more generic for future use. Modifiy Makefile and Kconfig
accordingly. Also, change driver compatible from "cdns,dp-phy" to
"cdns,torrent-phy".This will not affect ABI as the driver has never
been functional, and therefore do not exist in any active use case.
Signed-off-by: Yuti Amonkar <yamonkar@cadence.com>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
|
|
- Add Cadence MHDP PHY bindings in YAML format.
- Add Torrent PHY reference clock bindings.
- Add sub-node bindings for each group of PHY lanes based on PHY type.
Each sub-node includes properties such as master lane number, link reset,
phy type, number of lanes etc.
- Add reset support including PHY reset and individual lane reset.
- Add a new compatible string used for TI SoCs using Torrent PHY.
This will not affect ABI as the driver has never been functional,
and therefore do not exist in any active use case.
Signed-off-by: Yuti Amonkar <yamonkar@cadence.com>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
|
|
warning: `turbostat' uses 32-bit capabilities (legacy support in use)
Signed-off-by: Len Brown <len.brown@intel.com>
|
|
Some Chromebook BIOS' do not export an ACPI LPIT, which is how
Linux finds the residency counter for CPU and SYSTEM low power states,
that is exports in /sys/devices/system/cpu/cpuidle/*residency_us
When these sysfs attributes are missing, check the debugfs attrubte
from the pmc_core driver, which accesses the same counter value.
Signed-off-by: Len Brown <len.brown@intel.com>
|
|
From a turbostat point of view the Tremont-based Elkhart Lake
is very similar to Goldmont, reuse the code of Goldmont.
Elkhart Lake does not support 'group turbo limit counter'
nor C3, adjust the code accordingly.
Signed-off-by: Chen Yu <yu.c.chen@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
|
|
Jasper Lake, like Elkhart Lake, uses a Tremont CPU.
So reuse the code.
Signed-off-by: Chen Yu <yu.c.chen@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
|
|
From a turbostat point of view, Ice Lake server looks like Sky Lake server.
Signed-off-by: Chen Yu <yu.c.chen@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
|
|
From a turbostat point of view, Tiger Lake looks like Ice Lake.
Signed-off-by: Chen Yu <yu.c.chen@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
|
|
Warning: ‘__builtin_strncpy’ specified bound 20 equals destination size
[-Wstringop-truncation]
reduce param to strncpy, to guarantee that a null byte is always copied
into destination buffer.
Signed-off-by: Len Brown <len.brown@intel.com>
|
|
From a turbostat point of view, Cometlake is like Kabylake.
Suggested-by: Rui Zhang <rui.zhang@intel.com>
Signed-off-by: Chen Yu <yu.c.chen@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
|
|
git://anongit.freedesktop.org/drm/drm-intel into drm-fixes
drm/i915 fixes for v5.6-rc7:
- Track active elements during dequeue
- Fix failure to handle all MCR ranges
- Revert unnecessary workaround
Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/877dzgepvu.fsf@intel.com
|
|
git://people.freedesktop.org/~agd5f/linux into drm-fixes
amd-drm-fixes-5.6-2020-03-19:
amdgpu:
- Pageflip fix
- VCN clockgating fixes
- GPR debugfs fix for umr
- GPU reset fix
- eDP fix for MBP
- DCN2.x fix
Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Alex Deucher <alexdeucher@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200319204054.1036478-1-alexander.deucher@amd.com
|
|
With PR KVM, shutting down a VM causes the host kernel to crash:
[ 314.219284] BUG: Unable to handle kernel data access on read at 0xc00800000176c638
[ 314.219299] Faulting instruction address: 0xc008000000d4ddb0
cpu 0x0: Vector: 300 (Data Access) at [c00000036da077a0]
pc: c008000000d4ddb0: kvmppc_mmu_pte_flush_all+0x68/0xd0 [kvm_pr]
lr: c008000000d4dd94: kvmppc_mmu_pte_flush_all+0x4c/0xd0 [kvm_pr]
sp: c00000036da07a30
msr: 900000010280b033
dar: c00800000176c638
dsisr: 40000000
current = 0xc00000036d4c0000
paca = 0xc000000001a00000 irqmask: 0x03 irq_happened: 0x01
pid = 1992, comm = qemu-system-ppc
Linux version 5.6.0-master-gku+ (greg@palmb) (gcc version 7.5.0 (Ubuntu 7.5.0-3ubuntu1~18.04)) #17 SMP Wed Mar 18 13:49:29 CET 2020
enter ? for help
[c00000036da07ab0] c008000000d4fbe0 kvmppc_mmu_destroy_pr+0x28/0x60 [kvm_pr]
[c00000036da07ae0] c0080000009eab8c kvmppc_mmu_destroy+0x34/0x50 [kvm]
[c00000036da07b00] c0080000009e50c0 kvm_arch_vcpu_destroy+0x108/0x140 [kvm]
[c00000036da07b30] c0080000009d1b50 kvm_vcpu_destroy+0x28/0x80 [kvm]
[c00000036da07b60] c0080000009e4434 kvm_arch_destroy_vm+0xbc/0x190 [kvm]
[c00000036da07ba0] c0080000009d9c2c kvm_put_kvm+0x1d4/0x3f0 [kvm]
[c00000036da07c00] c0080000009da760 kvm_vm_release+0x38/0x60 [kvm]
[c00000036da07c30] c000000000420be0 __fput+0xe0/0x310
[c00000036da07c90] c0000000001747a0 task_work_run+0x150/0x1c0
[c00000036da07cf0] c00000000014896c do_exit+0x44c/0xd00
[c00000036da07dc0] c0000000001492f4 do_group_exit+0x64/0xd0
[c00000036da07e00] c000000000149384 sys_exit_group+0x24/0x30
[c00000036da07e20] c00000000000b9d0 system_call+0x5c/0x68
This is caused by a use-after-free in kvmppc_mmu_pte_flush_all()
which dereferences vcpu->arch.book3s which was previously freed by
kvmppc_core_vcpu_free_pr(). This happens because kvmppc_mmu_destroy()
is called after kvmppc_core_vcpu_free() since commit ff030fdf5573
("KVM: PPC: Move kvm_vcpu_init() invocation to common code").
The kvmppc_mmu_destroy() helper calls one of the following depending
on the KVM backend:
- kvmppc_mmu_destroy_hv() which does nothing (Book3s HV)
- kvmppc_mmu_destroy_pr() which undoes the effects of
kvmppc_mmu_init() (Book3s PR 32-bit)
- kvmppc_mmu_destroy_pr() which undoes the effects of
kvmppc_mmu_init() (Book3s PR 64-bit)
- kvmppc_mmu_destroy_e500() which does nothing (BookE e500/e500mc)
It turns out that this is only relevant to PR KVM actually. And both
32 and 64 backends need vcpu->arch.book3s to be valid when calling
kvmppc_mmu_destroy_pr(). So instead of calling kvmppc_mmu_destroy()
from kvm_arch_vcpu_destroy(), call kvmppc_mmu_destroy_pr() at the
beginning of kvmppc_core_vcpu_free_pr(). This is consistent with
kvmppc_mmu_init() being the last call in kvmppc_core_vcpu_create_pr().
For the same reason, if kvmppc_core_vcpu_create_pr() returns an
error then this means that kvmppc_mmu_init() was either not called
or failed, in which case kvmppc_mmu_destroy() should not be called.
Drop the line in the error path of kvm_arch_vcpu_create().
Fixes: ff030fdf5573 ("KVM: PPC: Move kvm_vcpu_init() invocation to common code")
Signed-off-by: Greg Kurz <groug@kaod.org>
Reviewed-by: Sean Christopherson <sean.j.christopherson@intel.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/158455341029.178873.15248663726399374882.stgit@bahia.lan
|
|
Commit '16f17eda8bad ("drm/amd/display: Send vblank and user
events at vsartup for DCN")' introduces a new way of pageflip
completion handling for DCN, and some trouble.
The current implementation introduces a race condition, which
can cause pageflip completion events to be sent out one vblank
too early, thereby confusing userspace and causing flicker:
prepare_flip_isr():
1. Pageflip programming takes the ddev->event_lock.
2. Sets acrtc->pflip_status == AMDGPU_FLIP_SUBMITTED
3. Releases ddev->event_lock.
--> Deadline for surface address regs double-buffering passes on
target pipe.
4. dc_commit_updates_for_stream() MMIO programs the new pageflip
into hw, but too late for current vblank.
=> pflip_status == AMDGPU_FLIP_SUBMITTED, but flip won't complete
in current vblank due to missing the double-buffering deadline
by a tiny bit.
5. VSTARTUP trigger point in vblank is reached, VSTARTUP irq fires,
dm_dcn_crtc_high_irq() gets called.
6. Detects pflip_status == AMDGPU_FLIP_SUBMITTED and assumes the
pageflip has been completed/will complete in this vblank and
sends out pageflip completion event to userspace and resets
pflip_status = AMDGPU_FLIP_NONE.
=> Flip completion event sent out one vblank too early.
This behaviour has been observed during my testing with measurement
hardware a couple of time.
The commit message says that the extra flip event code was added to
dm_dcn_crtc_high_irq() to prevent missing to send out pageflip events
in case the pflip irq doesn't fire, because the "DCH HUBP" component
is clock gated and doesn't fire pflip irqs in that state. Also that
this clock gating may happen if no planes are active. This suggests
that the problem addressed by that commit can't happen if planes
are active.
The proposed solution is therefore to only execute the extra pflip
completion code iff the count of active planes is zero and otherwise
leave pflip completion handling to the pflip irq handler, for a
more race-free experience.
Note that i don't know if this fixes the problem the original commit
tried to address, as i don't know what the test scenario was. It
does fix the observed too early pageflip events though and points
out the problem introduced.
Fixes: 16f17eda8bad ("drm/amd/display: Send vblank and user events at vsartup for DCN")
Reviewed-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com>
Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/ulfh/mmc
Pull MMC fixes from Ulf Hansson:
- rtsx_pci: Fix support for some various speed modes
- sdhci-of-at91: Fix support for GPIO card detect on SAMA5D2
- sdhci-cadence: Fix support for DDR52 speed mode for eMMC on UniPhier
- sdhci-acpi: Fix broken WP support on Acer Aspire Switch 10
- sdhci-acpi: Workaround FW bug for suspend on Lenovo Miix 320
* tag 'mmc-v5.6-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/ulfh/mmc:
mmc: rtsx_pci: Fix support for speed-modes that relies on tuning
mmc: sdhci-of-at91: fix cd-gpios for SAMA5D2
mmc: sdhci-cadence: set SDHCI_QUIRK2_PRESET_VALUE_BROKEN for UniPhier
mmc: sdhci-acpi: Disable write protect detection on Acer Aspire Switch 10 (SW5-012)
mmc: sdhci-acpi: Switch signal voltage back to 3.3V on suspend on external microSD on Lenovo Miix 320
|
